Haoyi Jiang
Haoyi Jiang
It is because that the `os.sep` under Windows is actually `\\` rather than `/` under Linux. The solution is simply replace the `os.sep` with `'/'` in the following 2 pieces...
#9 could be the possible cause
@ericmintun @HannaMao Sorry for bothering, but I think this issue is indeed serious and indisputable. Please pay attention to it. 🙏
#9 could be the possible cause, you may check the IoU scores for classes beyond 10.
> 1. Predict the depth from matched stereo images using the provided script with --dataset kitti360 --baseline 331.53255659999996 .(the baseline number was indicated in the script. Outputs look sensible) @ahayler...
@ahayler Thank you for your response. I have an additional question regarding: > Note: As stereo images were not provided in SSCBench and the pose files in the dataset also...
Did you load the trained checkpoint to generate outputs? We kept using the script to obtain the test score as reported. Since the train & val perforamce is right, I...
Assume it is solved since it haven't been updated for weeks
Thank you for bringing this to our attention! To clarify, I have modified my local code to evaluate the test set as the validation set for convenience. As a result,...
Hi, one more thing to confirm: have you adapted the disparity value for KITTI-360 as discussed [here](https://github.com/ai4ce/SSCBench/issues/8#issuecomment-1674607576)?